library(tidyverse)
library(gapminder)

\(\color{darkblue}{\textbf{Program}}\)

See Basic R

\(\color{darkblue}{\textbf{Import}}\)

Cheatsheet for Importing Data

Also see tidycensus and SQL

\(\color{dodgerblue}{\textbf{Equity Check}}\)

Where does this data come from?
Why was this data collected?
How was this data generated?
Is this data demographically representative?
Who is included and who is excluded from this data?
Whose voices, lives, and experiences are missing?
How much can this data be disaggregated by race, gender, ethnicity, etc.?
Are the categories mutually exclusive and fully inclusive?
Are there “other” categories and, if so, who does that include?
Who stands to benefit from this data?
Who might be harmed by the collection or publication of this data?

(See more in Urban Institute’s Do No Harm Guide)

\(\color{dodgerblue}{\textbf{Tibbles}}\)

as_tibble() transforms existing data into a tibble

data %>% 
  as_tibble()

tibble() creates a tibble from new data

tibble(x = 1:5, 
       y = 1, 
       z = x ^ 2 + y)

x	y	z
1	1	2
2	1	5
3	1	10
4	1	17
5	1	26

\(\color{dodgerblue}{\textbf{Import Data}}\)

\(\color{skyblue}{\textrm{- Basic Files (readr)}}\)

readr (part of tidyverse) covers basics like csv files

delim = ";" specifies the non-comma delimiter ("\t" is tab, " " is space)
skip = n skips the first n lines
comment = "#" drops all lines that start with (e.g.) #
col_names = FALSE tells read_csv() not to treat the first row as headings and instead label them from X1 to Xn
col_names = c("x", "y", "z") renames the columns as the character vector
col_types = cols() reads in the columns as the specified data type
na = 999 replaces dataset-specific NA values (e.g., 999) as NA

read_csv("path/to/file.csv")

read_rds("path/to/file.rds")

read_delim("path/to/file.txt", delim = ";")

#Best Practice
csv_data <- read_csv("path/to/file.csv",
                     col_types = cols(w = col_double(),
                                      x = col_date(format = ""),
                                      y = col_character(),
                                      z = col_logical()))

\(\color{skyblue}{\textrm{- Stats (haven)}}\)

haven reads SPSS, Stata, and SAS files

library(haven)

read_sas("path/to/file.sas7bdat")

read_spss("path/to/file.spv")

read_stata("path/to/file.dta")

\(\color{skyblue}{\textrm{- Excel (readxl)}}\)

readxl reads excel files (both .xls and .xlsx)

library(readxl)

read_excel("path/to/file.xlsx")

\(\color{skyblue}{\textrm{- Databases (DBI)}}\)

DBI, along with a database specific backend (e.g. RMySQL, RSQLite, RPostgreSQL etc), runs SQL queries (See SQL)

library(DBI)

con <- "connection to database"

dbGetQuery(con, "
  
  SELECT * 
    FROM table
           
           ") %>% as_tibble()

\(\color{skyblue}{\textrm{- JSON Files (jsonlite)}}\)

jsonlite reads json files

\(\color{skyblue}{\textrm{- XML Files (xml2)}}\)

xml2 reads XML

\(\color{dodgerblue}{\textbf{Export}}\)

write_csv() exports a plain csv

dataset %>% write_csv("path/dataset.csv")

write_rds() preserves formatting as an rds file

dataset %>% write_rds("path/dataset.rds")

\(\color{dodgerblue}{\textbf{Initial Check}}\)

\(\color{skyblue}{\textrm{- Contents}}\)

nrow() returns the number of rows

gapminder %>% nrow()

## [1] 1704

ncol() returns the number of columns

gapminder %>% ncol()

## [1] 6

names() returns a list of all of the columns

gapminder %>% names()

## [1] "country"   "continent" "year"      "lifeExp"   "pop"       "gdpPercap"

distinct() returns a list of all unique values in a specified column

gapminder %>% distinct(continent)

continent
Asia
Europe
Africa
Americas
Oceania

\(\color{skyblue}{\textrm{- Summary Stats}}\)

summary() gives basic stats on numeric columns, but I prefer this custom function

col_summaries <- function(.data){
  
  all_cols <- .data %>% names()
  
  test_col <- function(.test_col){
    test_data <- .data %>% 
      as_tibble() %>%
      rename(test = .test_col) %>% 
      mutate_if(is.logical, as.character)
    
    if(dim(test_data %>% filter(!is.na(test)))[1] == 0){
      test_data %>% 
        summarize(column = .test_col,
                  type = class(test_data$test),
                  missing = sum(is.na(test)),
                  unique_values = NA_real_,
                  example = NA_character_,
                  min = NA_real_,
                  q25 = NA_real_,
                  median = NA_real_,
                  q75 = NA_real_,
                  max = NA_real_)
    } else if(class(test_data$test) %in% c("integer", "numeric")){
      test_data %>% 
        summarize(column = .test_col,
                  type = class(test_data$test),
                  missing = sum(is.na(test)),
                  unique_values = NA_real_,
                  example = NA_character_,
                  min = min(test, na.rm = TRUE),
                  q25 = quantile(test, 0.25, na.rm = TRUE),
                  median = median(test, na.rm = TRUE),
                  q75 = quantile(test, 0.75, na.rm = TRUE),
                  max = max(test, na.rm = TRUE))
    } else {
      test_data %>% 
        summarize(column = .test_col,
                  type = class(test_data$test),
                  missing = sum(is.na(test)),
                  unique_values = n_distinct(test)) %>% 
        bind_cols(test_data %>% 
                    filter(!is.na(test)) %>% 
                    select(test) %>% 
                    head(1) %>% 
                    rename(example = test)) %>% 
        mutate(min = NA_real_,
               q25 = NA_real_,
               median = NA_real_,
               q75 = NA_real_,
               max = NA_real_ )
      } 
    }
  
  map_dfr(all_cols, test_col)
}

gapminder %>% col_summaries()

column	type	unique_values	example	min	q25	median	q75	max
country	factor	142	Afghanistan	NA	NA	NA	NA	NA
continent	factor	5	Asia	NA	NA	NA	NA	NA
year	integer	NA	NA	1952.0000	1965.750	1979.5000	1.993250e+03	2.007000e+03
lifeExp	numeric	NA	NA	23.5990	48.198	60.7125	7.084550e+01	8.260300e+01
pop	integer	NA	NA	60011.0000	2793664.000	7023595.5000	1.958522e+07	1.318683e+09
gdpPercap	numeric	NA	NA	241.1659	1202.060	3531.8470	9.325462e+03	1.135231e+05

\(\color{skyblue}{\textrm{- Missing Values}}\)

na = or recode sets what counts as missing

#When importing
dataframe <- read_csv("datasource.csv", 
                      na = c("", "NA", "UNKNOWN", 999))

#When already read-in
dateframe <- dataframe %>% 
  mutate(char_col = recode(char_col, "NA" = NA_character_),
         num_col = recode(num_col, 999 = NA_integer_))

any(is.na()) checks whether there are ANY missing values

#Overall
any(is.na(gapminder))

## [1] FALSE

#In a specific column
any(is.na(gapminder$lifeExp))

## [1] FALSE

summarize(sum(is.na())) counts the missing values within each column

gapminder %>% 
  summarize(observations = n(),
            missing_lifeExp = sum(is.na(lifeExp)))

observations	missing_lifeExp
1704	0

A list of expected values can also check completeness

expected <- c("Africa" = 54, 
              "Americas" = 35, 
              "Asia" = 47,
              "Europe" = 43,
              "Oceania" = 14)

gapminder %>% 
  filter(year == 2007) %>% 
  group_by(continent) %>% 
  summarize(countries_represented = n()) %>% 
  mutate(countries_expected = expected[continent],
         missing_countries = countries_expected - countries_represented)

continent	countries_represented	countries_expected	missing_countries
Africa	52	54	2
Americas	25	35	10
Asia	33	47	14
Europe	30	43	13
Oceania	2	14	12

\(\color{skyblue}{\textrm{- Cross Tabs}}\)

table() creates a cross tab count

gap_07 <- gapminder %>% 
  filter(year==2007) %>%
  mutate(country_size = case_when(pop > 3.121e+07 ~ "large",
                                  pop < 4.508e+06 ~ "small",
                                  TRUE            ~ "average"),
         country_size = factor(country_size, 
                               ordered = TRUE, 
                               levels = c("large", "average", "small")))

table(gap_07$continent, gap_07$country_size)

##           
##            large average small
##   Africa      10      25    17
##   Americas     6      13     6
##   Asia        13      14     6
##   Europe       7      17     6
##   Oceania      0       1     1

\(\color{skyblue}{\textrm{- Frequency Tables}}\)

prop.table() creates a frequency table, but I prefer this custom function for formatting

gap_07 <- gapminder %>% 
  filter(year==2007) %>%
  mutate(country_size = case_when(pop > 3.121e+07 ~ "large",
                                  pop < 4.508e+06 ~ "small",
                                  TRUE            ~ "average"),
         country_size = factor(country_size, 
                               ordered = TRUE, 
                               levels = c("large", "average", "small")))

size_table <- table(gap_07$continent, gap_07$country_size)

props <- function(col1, col2, prop_type){
  cross_tab <- table(col1, col2) 
  
  if(prop_type == "overall"){
    prop.table(cross_tab) %>% 
      as.data.frame() %>% 
      as_tibble() %>% 
      spread(col2, Freq) %>%
      janitor::adorn_totals("col") %>% 
      janitor::adorn_totals("row") %>% 
      as_tibble() %>% 
      mutate_at(vars(-col1), scales::percent, accuracy = 0.1) %>% 
      rename(` ` = col1)
  } else if(prop_type == "row"){
    prop.table(cross_tab, 2) %>% 
      as.data.frame() %>% 
      as_tibble() %>% 
      spread(col2, Freq) %>%
      janitor::adorn_totals("row") %>% 
      as_tibble() %>% 
      mutate_at(vars(-col1), scales::percent, accuracy = 0.1) %>% 
      rename(` ` = col1)
  } else{
    prop.table(cross_tab, 1) %>% 
      as.data.frame() %>% 
      as_tibble() %>% 
      spread(col2, Freq) %>%
      janitor::adorn_totals("col") %>% 
      as_tibble() %>% 
      mutate_at(vars(-col1), scales::percent, accuracy = 0.1) %>% 
      rename(` ` = col1)
  }
}

#Overall Proportions
props(gap_07$continent, gap_07$country_size, "overall")

	large	average	small	Total
Africa	7.0%	17.6%	12.0%	36.6%
Americas	4.2%	9.2%	4.2%	17.6%
Asia	9.2%	9.9%	4.2%	23.2%
Europe	4.9%	12.0%	4.2%	21.1%
Oceania	0.0%	0.7%	0.7%	1.4%
Total	25.4%	49.3%	25.4%	100.0%

#Column Proportions
props(gap_07$continent, gap_07$country_size, "col")

	large	average	small	Total
Africa	19.2%	48.1%	32.7%	100.0%
Americas	24.0%	52.0%	24.0%	100.0%
Asia	39.4%	42.4%	18.2%	100.0%
Europe	23.3%	56.7%	20.0%	100.0%
Oceania	0.0%	50.0%	50.0%	100.0%

#Row Proportions
props(gap_07$continent, gap_07$country_size, "row")

	large	average	small
Africa	27.8%	35.7%	47.2%
Americas	16.7%	18.6%	16.7%
Asia	36.1%	20.0%	16.7%
Europe	19.4%	24.3%	16.7%
Oceania	0.0%	1.4%	2.8%
Total	100.0%	100.0%	100.0%

\(\color{skyblue}{\textrm{- Correlation Tables}}\)

cor() calculated the correlation between two variables, but this function will compare each variable to all other variables

library(reshape2)

get_pretty_cormap <- function (.data, .numcols) {
  cormat <- round(cor(select(.data, .numcols)), 2)
  
  reorder_cormat <- function(cormat){
    dd <- as.dist((1-cormat)/2)
    hc <- hclust(dd)
    cormat <-cormat[hc$order, hc$order]}
  cormat <- reorder_cormat(cormat)
  
  get_upper_tri <- function(cormat){
    cormat[lower.tri(cormat)]<- NA
    return(cormat)}
  upper_tri <- get_upper_tri(cormat)
  
  melted_cormat <- melt(upper_tri, na.rm = TRUE)
  
  ggplot(melted_cormat, aes(Var2, Var1, fill = value)) +
    geom_tile(color = "white") +
    scale_fill_gradient2(low = "blue", high = "red", mid = "white", midpoint = 0, limit = c(-1,1), 
                         space = "Lab", name = "Pearson\nCorrelation") +
    theme_minimal() +
    theme(axis.text.x = element_text(angle = 45, vjust = 1, size = 12, hjust = 1)) +
    coord_fixed() +
    geom_text(aes(Var2, Var1, label = value), color = "black", size = 4) +
    theme(axis.title.x = element_blank(),
          axis.title.y = element_blank(),
          panel.grid.major = element_blank(),
          panel.border = element_blank(),
          panel.background = element_blank(),
          axis.ticks = element_blank(),
          legend.justification = c(1, 0),
          legend.position = c(0.6, 0.7),   
          legend.direction = "horizontal") +
    guides(fill = guide_colorbar(barwidth = 7, barheight = 1, title.position = "top", title.hjust = 0.5))}

get_pretty_cormap(gapminder, c("year", "lifeExp", "pop", "gdpPercap"))

\(\color{darkblue}{\textbf{Tidy}}\)

Cheatsheet for Tidying Data

Cheatsheet for Strings

Cheatsheet for Factors

Cheatsheet for Lubridates

\(\color{dodgerblue}{\textbf{Tidy Data}}\)

\(\color{skyblue}{\textrm{- Gather}}\)

dataframe %>% 
  gather(-c(cols to keep),
         key = key name, 
         value = value name, 
         factor_key = TRUE)

ratings %>% 
  gather(-Season, 
         key = Episode, 
         value = Rating, 
         factor_key = TRUE)

\(\color{skyblue}{\textrm{- Spread}}\)

dataframe %>% 
  spread(key = col to become new cols, 
         value = col to become values under new cols, 
         fill = value to fill for missing)

roster %>% 
  spread(key = Variable, 
         value = Rating)

\(\color{dodgerblue}{\textbf{Values}}\)

\(\color{skyblue}{\textrm{- Missing}}\)

filter(!is.na()) removes entire rows with missing values in the specified column

gapminder %>% 
  filter(!is.na(lifeExp))

replace_na() replaces missing values

gapminder %>% 
  mutate(pop = replace_na(pop, 0))

\(\color{skyblue}{\textrm{- Duplicates}}\)

distinct() keeps unique rows and removes all others that are completely identical

data %>% 
  distinct()

distinct(col, .keep_all = TRUE) (i.e., adding a column as an argument) just looks at the values within that column regardless of whether the other columns are the same (this will keep the first occurrence, so arrange as desired first)

#Most recent entry for each country
gapminder %>% 
  arrange(desc(year)) %>% 
  distinct(country, .keep_all = TRUE) %>% 
  head()

country	continent	year	lifeExp	pop	gdpPercap
Afghanistan	Asia	2007	43.828	31889923	974.5803
Albania	Europe	2007	76.423	3600523	5937.0295
Algeria	Africa	2007	72.301	33333216	6223.3675
Angola	Africa	2007	42.731	12420476	4797.2313
Argentina	Americas	2007	75.320	40301927	12779.3796
Australia	Oceania	2007	81.235	20434176	34435.3674

\(\color{skyblue}{\textrm{- Outliers}}\)

Causes:

Valid measurements that happen to be extreme
Variability in measurement
Experimental error
Data entry error

#Visual Check
boxplot(gap_07$pop)

All of the circles are technically outliers. We know that the two extremes (China and India) are, in fact, correct. If there was a third that was that extreme, it would be an error that should be removed or replaced. Similarly, if there was a negative population, that would also be an error that should be removed or replaced.

#3SD Check
upper_out <- mean(gap_07$pop) + (3 * sd(gap_07$pop))
lower_out <- mean(gap_07$pop) - (3 * sd(gap_07$pop))

gap_07 %>% 
  filter(pop > upper_out | pop < lower_out)

country	continent	year	lifeExp	pop	gdpPercap	country_size
China	Asia	2007	72.961	1318683096	4959.115	large
India	Asia	2007	64.698	1110396331	2452.210	large

\(\color{dodgerblue}{\textbf{Selecting}}\)

\(\color{skyblue}{\textrm{- Filter (observations)}}\)

filter() selects all observations (rows) with values matching the specification(s)

Code	Meaning
`==`	Equal to
`!=`	Not equal to
`>`	Greater than
`>=`	Greater than or equal to
`<`	Less than
`<=`	Less than or equal to
`%in% c()`	Within the character vector
`!(. %in% c())`	Not within the character vector
`is.na()`	Is NA
`!is.na()`	Is not NA
`between(x, y)`	Is between x and y
`,` or `&`	And
`\|`	Or

gapminder %>% 
  filter(between(year, 2000, 2020),
         !(continent %in% c("Africa", "Americas", "Asia")),
         !is.na(gdpPercap)) %>% 
  filter(lifeExp > 81 |
         gdpPercap > 40000) %>% 
  head()

country	continent	year	lifeExp	pop	gdpPercap
Australia	Oceania	2007	81.235	20434176	34435.37
Iceland	Europe	2007	81.757	301931	36180.79
Ireland	Europe	2007	78.885	4109086	40676.00
Norway	Europe	2002	79.050	4535591	44683.98
Norway	Europe	2007	80.196	4627926	49357.19
Switzerland	Europe	2007	81.701	7554661	37506.42

\(\color{skyblue}{\textrm{- Select (variables)}}\)

select() selects all variables (columns) called for

starts_with("abc") selects columns that begin with “abc”
ends_with("xyz") selects columns that end with “xyz”
contains("ijk") selects columns that contain “ijk”
matches("(.)\\1") selects columns that match a regular expression
everything() selects everything else (useful for reorganizing columns)

#Just
gapminder %>% 
  select(country, year, pop) %>% 
  head()

country	year	pop
Afghanistan	1952	8425333
Afghanistan	1957	9240934
Afghanistan	1962	10267083
Afghanistan	1967	11537966
Afghanistan	1972	13079460
Afghanistan	1977	14880372

#Not
gapminder %>% 
  select(-c(continent, pop)) %>% 
  head()

country	year	lifeExp	gdpPercap
Afghanistan	1952	28.801	779.4453
Afghanistan	1957	30.332	820.8530
Afghanistan	1962	31.997	853.1007
Afghanistan	1967	34.020	836.1971
Afghanistan	1972	36.088	739.9811
Afghanistan	1977	38.438	786.1134

#Range from this to that
gapminder %>% 
  select(continent:pop) %>% 
  head()

continent	year	lifeExp	pop
Asia	1952	28.801	8425333
Asia	1957	30.332	9240934
Asia	1962	31.997	10267083
Asia	1967	34.020	11537966
Asia	1972	36.088	13079460
Asia	1977	38.438	14880372

\(\color{dodgerblue}{\textbf{Columns}}\)

\(\color{skyblue}{\textrm{- Renaming}}\)

rename(new name = old name) renames the column

gapminder %>%
  rename(population = pop) %>% 
  head()

country	continent	year	lifeExp	population	gdpPercap
Afghanistan	Asia	1952	28.801	8425333	779.4453
Afghanistan	Asia	1957	30.332	9240934	820.8530
Afghanistan	Asia	1962	31.997	10267083	853.1007
Afghanistan	Asia	1967	34.020	11537966	836.1971
Afghanistan	Asia	1972	36.088	13079460	739.9811
Afghanistan	Asia	1977	38.438	14880372	786.1134

\(\color{skyblue}{\textrm{- Separating}}\)

separate() splits each value within a column based on some separator

gapminder %>% 
  separate(col = lifeExp, 
           into = c("Years", "Months"), 
           sep = "[.]") %>%
  mutate(Months = as.double(str_c("0.", Months)) * 12) %>% 
  head()

country	continent	year	Years	Months	pop	gdpPercap
Afghanistan	Asia	1952	28	9.612	8425333	779.4453
Afghanistan	Asia	1957	30	3.984	9240934	820.8530
Afghanistan	Asia	1962	31	11.964	10267083	853.1007
Afghanistan	Asia	1967	34	0.240	11537966	836.1971
Afghanistan	Asia	1972	36	1.056	13079460	739.9811
Afghanistan	Asia	1977	38	5.256	14880372	786.1134

\(\color{skyblue}{\textrm{- Uniting}}\)

unite() merges the values of two columns into a new column, placing some separator between the values

gapminder %>% 
  unite(col = location, 
        country, continent, 
        sep = ", ") %>%
  head()

location	year	lifeExp	pop	gdpPercap
Afghanistan, Asia	1952	28.801	8425333	779.4453
Afghanistan, Asia	1957	30.332	9240934	820.8530
Afghanistan, Asia	1962	31.997	10267083	853.1007
Afghanistan, Asia	1967	34.020	11537966	836.1971
Afghanistan, Asia	1972	36.088	13079460	739.9811
Afghanistan, Asia	1977	38.438	14880372	786.1134

\(\color{dodgerblue}{\textbf{Mutating}}\)

mutate() adds columns to the existing dataframe

gapminder %>% 
  mutate(gdp = gdpPercap * as.double(pop)) %>% 
  head()

country	continent	year	lifeExp	pop	gdpPercap	gdp
Afghanistan	Asia	1952	28.801	8425333	779.4453	6567086330
Afghanistan	Asia	1957	30.332	9240934	820.8530	7585448670
Afghanistan	Asia	1962	31.997	10267083	853.1007	8758855797
Afghanistan	Asia	1967	34.020	11537966	836.1971	9648014150
Afghanistan	Asia	1972	36.088	13079460	739.9811	9678553274
Afghanistan	Asia	1977	38.438	14880372	786.1134	11697659231

transmute() adds columns and then only selects the specified columns

gapminder %>% 
  transmute(country, 
            year, 
            gdp = gdpPercap * as.double(pop)) %>% 
  head()

country	year	gdp
Afghanistan	1952	6567086330
Afghanistan	1957	7585448670
Afghanistan	1962	8758855797
Afghanistan	1967	9648014150
Afghanistan	1972	9678553274
Afghanistan	1977	11697659231

\(\color{skyblue}{\textrm{- Recasting}}\)

as.character()
as.double()
as.integer()
as.logical()

gapminder %>% 
  mutate(pop = as.double(pop)) %>% 
  head()

country	continent	year	lifeExp	pop	gdpPercap
Afghanistan	Asia	1952	28.801	8425333	779.4453
Afghanistan	Asia	1957	30.332	9240934	820.8530
Afghanistan	Asia	1962	31.997	10267083	853.1007
Afghanistan	Asia	1967	34.020	11537966	836.1971
Afghanistan	Asia	1972	36.088	13079460	739.9811
Afghanistan	Asia	1977	38.438	14880372	786.1134

\(\color{skyblue}{\textrm{- Factoring}}\)

factor() recategorizes strings as categorical data; ordered = TRUE orders the categories by the levels = c() argument (otherwise it will be alphabetical)

gapminder %>% 
  mutate(country = factor(country),
         continent = factor(continent, 
                            ordered = TRUE,
                            levels = c("Asia", "Americas", "Europe", 
                                       "Africa", "Oceania"))) %>% 
  head()

country	continent	year	lifeExp	pop	gdpPercap
Afghanistan	Asia	1952	28.801	8425333	779.4453
Afghanistan	Asia	1957	30.332	9240934	820.8530
Afghanistan	Asia	1962	31.997	10267083	853.1007
Afghanistan	Asia	1967	34.020	11537966	836.1971
Afghanistan	Asia	1972	36.088	13079460	739.9811
Afghanistan	Asia	1977	38.438	14880372	786.1134

\(\color{skyblue}{\textrm{- Cases}}\)

case_when() creates a new field of categorical data (that still needs to be factored)

 gapminder %>% 
  filter(year == 2007) %>%
  mutate(country_size = case_when(pop > 3.121e+07 ~ "Large",
                                  pop < 4.508e+06 ~ "Small",
                                  TRUE            ~ "Average"),
         country_size = factor(country_size,
                               ordered = TRUE,
                               levels = c("Small", "Average", "Large"))) %>% 
  head()

country	continent	year	lifeExp	pop	gdpPercap	country_size
Afghanistan	Asia	2007	43.828	31889923	974.5803	Large
Albania	Europe	2007	76.423	3600523	5937.0295	Small
Algeria	Africa	2007	72.301	33333216	6223.3675	Large
Angola	Africa	2007	42.731	12420476	4797.2313	Average
Argentina	Americas	2007	75.320	40301927	12779.3796	Large
Australia	Oceania	2007	81.235	20434176	34435.3674	Average

\(\color{skyblue}{\textrm{- Recoding Values}}\)

recode(old value = new value) replaces all occurrences of the old value with the new value

 gapminder %>% 
  mutate(continent = recode(continent, 
                            "Asia" = "Asia & Pacific Islands",
                            "Oceania" = "Australia & New Zealand")) %>% 
  head()

country	continent	year	lifeExp	pop	gdpPercap
Afghanistan	Asia & Pacific Islands	1952	28.801	8425333	779.4453
Afghanistan	Asia & Pacific Islands	1957	30.332	9240934	820.8530
Afghanistan	Asia & Pacific Islands	1962	31.997	10267083	853.1007
Afghanistan	Asia & Pacific Islands	1967	34.020	11537966	836.1971
Afghanistan	Asia & Pacific Islands	1972	36.088	13079460	739.9811
Afghanistan	Asia & Pacific Islands	1977	38.438	14880372	786.1134

A lookup table recodes all values faster:

library(hflights)

hflights %>% 
  count(UniqueCarrier) %>% 
  arrange(desc(n)) %>% 
  head()

UniqueCarrier	n
XE	73053
CO	70032
WN	45343
OO	16061
MQ	4648
US	4082

airline_lookup_table <- c("AA" = "American",
                          "AS" = "Alaska",
                          "B6" = "JetBlue",
                          "CO" = "Continental",
                          "DL" = "Delta",
                          "OO" = "SkyWest",
                          "UA" = "United",
                          "US" = "US_Airways", 
                          "WN" = "Southwest", 
                          "EV" = "Atlantic_Southeast", 
                          "F9" = "Frontier", 
                          "FL" = "AirTran", 
                          "MQ" = "American_Eagle", 
                          "XE" = "ExpressJet", 
                          "YV" = "Mesa")

hflights %>% 
  mutate(UniqueCarrier = airline_lookup_table[UniqueCarrier]) %>%
  count(UniqueCarrier) %>% 
  arrange(desc(n)) %>% 
  head()

UniqueCarrier	n
ExpressJet	73053
Continental	70032
Southwest	45343
SkyWest	16061
American_Eagle	4648
US_Airways	4082

\(\color{skyblue}{\textrm{- Leading/Lagging Values}}\)

First order with arrange() and arrange(desc()), then lead() and lag() pull the leading or lagging value

gapminder %>% 
  filter(country == "United States",
         year > 1990) %>% 
  arrange(year) %>% 
  transmute(year,
            gdpPercap,
            prior_gdpPercap = lag(gdpPercap),
            next_gdpPercap = lead(gdpPercap))

year	gdpPercap	prior_gdpPercap	next_gdpPercap
1992	32003.93	NA	35767.43
1997	35767.43	32003.93	39097.10
2002	39097.10	35767.43	42951.65
2007	42951.65	39097.10	NA

\(\color{skyblue}{\textrm{- Rolling Analysis}}\)

First order with arrange() and arrange(desc()), then cum_() does a rolling analysis

gapminder %>% 
  filter(country == "United States",
         year > 1990) %>% 
  arrange(year) %>% 
  transmute(year,
            gdpPercap,
            rolling_sum = cumsum(gdpPercap), 
            rolling_mean = cummean(gdpPercap), 
            min_to_date = cummin(gdpPercap),
            max_to_date = cummax(gdpPercap))

year	gdpPercap	rolling_sum	rolling_mean	min_to_date	max_to_date
1992	32003.93	32003.93	32003.93	32003.93	32003.93
1997	35767.43	67771.37	33885.68	32003.93	35767.43
2002	39097.10	106868.46	35622.82	32003.93	39097.10
2007	42951.65	149820.12	37455.03	32003.93	42951.65

\(\color{skyblue}{\textrm{- Whole Column Analysis}}\)

Calling a column by name analyzes the whole column

gapminder %>% 
  filter(year == 2007) %>% 
  transmute(country,
            continent,
            lifeExp,
            world_lifeExp = mean(lifeExp, na.rm = TRUE),
            above_world_lifeExp = lifeExp > world_lifeExp,
            pop) %>% 
  arrange(desc(pop)) %>% 
  head()

country	continent	lifeExp	world_lifeExp	above_world_lifeExp	pop
China	Asia	72.961	67.00742	TRUE	1318683096
India	Asia	64.698	67.00742	FALSE	1110396331
United States	Americas	78.242	67.00742	TRUE	301139947
Indonesia	Asia	70.650	67.00742	TRUE	223547000
Brazil	Americas	72.390	67.00742	TRUE	190010647
Pakistan	Asia	65.483	67.00742	FALSE	169270617

\(\color{skyblue}{\textrm{- Rankings}}\)

ntile() categorizes the data into n even groups
rank() ranks the column

gapminder %>% 
  filter(year == 2007) %>% 
  transmute(country,
            continent,
            lifeExp,
            lifeExpQuantile = ntile(desc(lifeExp), 4),
            lifeExpRank = rank(desc(lifeExp)),
            pop) %>% 
  arrange(desc(pop)) %>% 
  head()

country	continent	lifeExp	lifeExpQuantile	lifeExpRank	pop
China	Asia	72.961	2	59	1318683096
India	Asia	64.698	3	91	1110396331
United States	Americas	78.242	1	30	301139947
Indonesia	Asia	70.650	3	80	223547000
Brazil	Americas	72.390	2	68	190010647
Pakistan	Asia	65.483	3	89	169270617

\(\color{skyblue}{\textrm{- Totals}}\)

The janitor package can adorn_totals()

#Row total
gapminder %>% 
  filter(continent == "Oceania",
         year > 1990) %>% 
  select(country, year, pop) %>% 
  spread(year, pop) %>% 
  janitor::adorn_totals("col") %>% 
  as_tibble()

country	1992	1997	2002	2007	Total
Australia	17481977	18565243	19546792	20434176	76028188
New Zealand	3437674	3676187	3908037	4115771	15137669

#Column total
gapminder %>% 
  filter(continent == "Oceania",
         year > 1990) %>% 
  select(country, year, pop) %>% 
  spread(year, pop) %>% 
  janitor::adorn_totals("row") %>% 
  as_tibble()

country	1992	1997	2002	2007
Australia	17481977	18565243	19546792	20434176
New Zealand	3437674	3676187	3908037	4115771
Total	20919651	22241430	23454829	24549947

#Both totals
gapminder %>% 
  filter(continent == "Oceania",
         year > 1990) %>% 
  select(country, year, pop) %>% 
  spread(year, pop) %>% 
  janitor::adorn_totals("row") %>%
  janitor::adorn_totals("col") %>%
  as_tibble()

country	1992	1997	2002	2007	Total
Australia	17481977	18565243	19546792	20434176	76028188
New Zealand	3437674	3676187	3908037	4115771	15137669
Total	20919651	22241430	23454829	24549947	91165857

\(\color{skyblue}{\textrm{- Grouped Analysis}}\)

group_by can perform any analysis within a group instead of over the whole column

gapminder %>% 
  filter(year == 2007) %>%
  transmute(country,
            continent,
            lifeExp,
            overall_quartile = ntile(desc(lifeExp), 4)) %>% 
  group_by(continent) %>% 
  mutate(continent_quantile = ntile(desc(lifeExp), 4)) %>% 
  ungroup() %>% 
  arrange(continent) %>% 
  head()

country	continent	lifeExp	overall_quartile	continent_quantile
Algeria	Africa	72.301	2	1
Angola	Africa	42.731	4	4
Benin	Africa	56.728	4	2
Botswana	Africa	50.728	4	3
Burkina Faso	Africa	52.295	4	3
Burundi	Africa	49.580	4	3

\(\color{skyblue}{\textrm{- Multiple Columns}}\)

mutate_all, mutate_at, and mutate_if analyze multiple columns simultaneously with the same function

#All 
gapminder %>% 
  mutate_all(as.character) %>% 
  head()

country	continent	year	lifeExp	pop	gdpPercap
Afghanistan	Asia	1952	28.801	8425333	779.4453145
Afghanistan	Asia	1957	30.332	9240934	820.8530296
Afghanistan	Asia	1962	31.997	10267083	853.10071
Afghanistan	Asia	1967	34.02	11537966	836.1971382
Afghanistan	Asia	1972	36.088	13079460	739.9811058
Afghanistan	Asia	1977	38.438	14880372	786.11336

#Just
gapminder %>% 
  mutate_at(vars(lifeExp, gdpPercap), round, 1) %>% 
  head()

country	continent	year	lifeExp	pop	gdpPercap
Afghanistan	Asia	1952	28.8	8425333	779.4
Afghanistan	Asia	1957	30.3	9240934	820.9
Afghanistan	Asia	1962	32.0	10267083	853.1
Afghanistan	Asia	1967	34.0	11537966	836.2
Afghanistan	Asia	1972	36.1	13079460	740.0
Afghanistan	Asia	1977	38.4	14880372	786.1

#All but
gapminder %>% 
  mutate_at(vars(-country, -continent), as.double) %>% 
  head()

country	continent	year	lifeExp	pop	gdpPercap
Afghanistan	Asia	1952	28.801	8425333	779.4453
Afghanistan	Asia	1957	30.332	9240934	820.8530
Afghanistan	Asia	1962	31.997	10267083	853.1007
Afghanistan	Asia	1967	34.020	11537966	836.1971
Afghanistan	Asia	1972	36.088	13079460	739.9811
Afghanistan	Asia	1977	38.438	14880372	786.1134

#Like
gapminder %>% 
  mutate_at(vars(matches("gdp")), round, 1) %>% 
  head()

country	continent	year	lifeExp	pop	gdpPercap
Afghanistan	Asia	1952	28.801	8425333	779.4
Afghanistan	Asia	1957	30.332	9240934	820.9
Afghanistan	Asia	1962	31.997	10267083	853.1
Afghanistan	Asia	1967	34.020	11537966	836.2
Afghanistan	Asia	1972	36.088	13079460	740.0
Afghanistan	Asia	1977	38.438	14880372	786.1

#If
gapminder %>% 
  mutate_if(is.numeric, scales::comma, accuracy = 0.1) %>% 
  head()

country	continent	year	lifeExp	pop	gdpPercap
Afghanistan	Asia	1,952.0	28.8	8,425,333.0	779.4
Afghanistan	Asia	1,957.0	30.3	9,240,934.0	820.9
Afghanistan	Asia	1,962.0	32.0	10,267,083.0	853.1
Afghanistan	Asia	1,967.0	34.0	11,537,966.0	836.2
Afghanistan	Asia	1,972.0	36.1	13,079,460.0	740.0
Afghanistan	Asia	1,977.0	38.4	14,880,372.0	786.1

\(\color{dodgerblue}{\textbf{Summarizing}}\)

summarize() groups and returns a single summary statistic

gapminder %>% 
  filter(year == 2007) %>% 
  group_by(continent) %>% 
  summarize(count = n(), 
            missing = sum(is.na(lifeExp)),
            unique_countries = n_distinct(country),
            low_lifeExp = min(lifeExp, na.rm = TRUE),
            high_lifeExp = max(lifeExp, na.rm = TRUE),
            mean_lifeExp = mean(lifeExp, na.rm = TRUE),
            median_lifeEx = median(lifeExp, na.rm = TRUE),
            above_world_mean = sum(lifeExp > mean(gapminder %>% 
                                                    filter(year == 2007) %>% 
                                                    .$lifeExp,
                                                  na.rm = TRUE)),
            share_above_world_mean = above_world_mean / count) %>% 
  ungroup()

continent	count	unique_countries	low_lifeExp	high_lifeExp	mean_lifeExp	median_lifeEx	above_world_mean	share_above_world_mean
Africa	52	52	39.613	76.442	54.80604	52.9265	7	0.1346154
Americas	25	25	60.916	80.653	73.60812	72.8990	23	0.9200000
Asia	33	33	43.828	82.603	70.72848	72.3960	23	0.6969697
Europe	30	30	71.777	81.757	77.64860	78.6085	30	1.0000000
Oceania	2	2	80.204	81.235	80.71950	80.7195	2	1.0000000

\(\color{dodgerblue}{\textbf{Strings}}\)

\(\color{skyblue}{\textrm{- Trimming Spaces}}\)

str_trim() cuts out spaces at the beginning and end of the string

str_trim("  this is a test     ")

## [1] "this is a test"

\(\color{skyblue}{\textrm{- Padding}}\)

str_pad() adds a number (width) of characters (pad) at the beginning (left) or end (right) of a string

county_codes <- c("1", "51", "201")

str_pad(county_codes, 
        width = 3, 
        side = "left", 
        pad = "0")

## [1] "001" "051" "201"

\(\color{skyblue}{\textrm{- Extract Number}}\)

parse_number() drops all non-numeric characters

parse_number("$1,234,567")

## [1] 1234567

\(\color{skyblue}{\textrm{- Replacing}}\)

str_replace(string, old, new) swaps out old for new characters

#Replace First
str_replace("bananas","a","o")

## [1] "bonanas"

#Replace All
str_replace_all("bananas","a","o")

## [1] "bononos"

str_remove(string, old) just drops the old characters

#Remove  First
str_remove("bananas","a")

## [1] "bnanas"

#Remove All
str_remove_all("bananas","a")

## [1] "bnns"

\(\color{skyblue}{\textrm{- Capitalization}}\)

str has some capitalization conventions (though setting all to lower is best for matching)

text <- "nEw YoRk CiTy"

#All Lower
str_to_lower(text)

## [1] "new york city"

#All Upper
str_to_upper(text)

## [1] "NEW YORK CITY"

#First Word Capitalized
str_to_sentence(text)

## [1] "New york city"

#All Words Capitalized
str_to_title(text)

## [1] "New York City"

\(\color{dodgerblue}{\textbf{Dates}}\)

library(lubridate)

A date-time is a point on the timeline stored as the number of sections since 1970-01-01 00:00:00 UTC

as_datetime(1511870400)

## [1] "2017-11-28 12:00:00 UTC"

A date is a point on the timeline stored as the number of days since 1970-01-01

as_date(17498)

## [1] "2017-11-28"

A time is a point on the timeline stored as the number of seconds since 00:00:00

hms::as_hms(10230)

## 02:50:30

\(\color{skyblue}{\textrm{- Current}}\)

now() returns the date-time

now()

## [1] "2022-02-10 17:46:04 EST"

today() returns the date

today()

## [1] "2022-02-10"

\(\color{skyblue}{\textrm{- Parsing}}\)

The functions match the order of the components. For example, 08/07/06 could be:

Aug 7, 2006
Jul 8, 2006
Jul 6, 2008
Jun 7, 2008

mdy("08/07/06")

## [1] "2006-08-07"

dmy("08/07/06")

## [1] "2006-07-08"

ymd("08/07/06")

## [1] "2008-07-06"

ydm("08/07/06")

## [1] "2008-06-07"

It also works for more complicated renderings like “The 4th of July, 2018”

dmy("The 4th of July, 2018")

## [1] "2018-07-04"

And for time

mdy_hms("08/07/06 10:09:30")

## [1] "2006-08-07 10:09:30 UTC"

For more complex strings, the format can be specified:

y = year (2021 or 21, but defaults to current century)
m = numerical month (1-12 or 01-12)
b = month name (January or Jan)
d = day of the month (1-31 or 01-31)
w = numerical day of the week (0-6 with Sunday as 0)
a = day of the week name (Monday or Mon)
_ = split between date and time
H = 24 hour (0-24 or 00-24)
I = am/pm hour (1-12 or 01-12)
p = am or pm when using I
M = minute (0-59 or 00-59)
s = second (0-61 or 00-61)

parse_date_time("Monday June 1st 2010 at 4pm", orders = "amdy_Ip")

## [1] "2010-06-01 16:00:00 UTC"

\(\color{skyblue}{\textrm{- Time Zones}}\)

tz() checks the timezone

tz(ymd_hms("2017-03-11 12:00:00"))

## [1] "UTC"

tz = sets the timezone

ymd_hms("2017-03-11 12:00:00", tz = "America/Los_Angeles")

## [1] "2017-03-11 12:00:00 PST"

force_tz() changes an incorrect timezone (without changing the time value)

force_tz(ymd_hms("2017-03-11 12:00:00"), tzone = "America/New_York")

## [1] "2017-03-11 12:00:00 EST"

with_tz converts to a new timezone (thus changing the time value)

with_tz(ymd_hms("2017-03-11 12:00:00"), tzone = "America/New_York")

## [1] "2017-03-11 07:00:00 EST"

\(\color{skyblue}{\textrm{- Extracting Elements}}\)

date() = drops the time
year() = year
month() = month
day() = number day of the month
yday() = number day of the year (1-365)
wday() = day of the week
hour() = hour
min() = minute
leap_year() = true/false for leap year
dst() = true/false for daylight savings
quarter() = quarter of year
semester() = half of year
label = TRUE for word instead of number
abbr = TRUE for abbreviated word

str_c("Today is day ", day(today()),
      " of the month of ", month(today(), label = TRUE, abbr = FALSE),
      ", which is a ", wday(today(), label = TRUE, abbr = FALSE),
      ". It is also day ", yday(today()), 
      " of ", year(today()),
      ".")

## [1] "Today is day 10 of the month of February, which is a Thursday. It is also day 41 of 2022."

\(\color{skyblue}{\textrm{- Rounding}}\)

round_date() gives the nearest unit

round_date(ymd("2021-06-30"), unit = "month")

## [1] "2021-07-01"

ceiling_date() rounds up

ceiling_date(ymd_hm("2021-06-30 12:50"), unit = "15 minutes")

## [1] "2021-06-30 13:00:00 UTC"

floor_date() rounds down

floor_date(ymd_hm("2021-06-30 12:50"), unit = "hour")

## [1] "2021-06-30 12:00:00 UTC"

\(\color{skyblue}{\textrm{- Adding Time}}\)

Time does not behave like general number lines. To evaluate, you must ask which is more important: the datetime in the real world or the length of time (duration)?

seconds() / dseconds()
minutes() / dminutes()
hours() / dhours()
days() / ddays()
weeks() / dweeks()
months() / dmonths()
years() / dyears()

#Normal Day
normal_day <- ymd_hm("2018-01-01 24:00", tz = "America/New_York")

period <- normal_day + hours(3)

duration <- normal_day + dhours(3)

tibble(Difference = difftime(period, duration, units = "hours"),
       Period = period,
       Duration = duration)

Difference	Period	Duration
0 hours	2018-01-02 03:00:00	2018-01-02 03:00:00

#Daylight Savings
daylight_savings_start <- ymd_hm("2018-03-10 24:00", tz = "America/New_York")

period <- daylight_savings_start + hours(3)

duration <- daylight_savings_start + dhours(3)

tibble(Difference = difftime(period, duration, units = "hours"),
       Period = period,
       Duration = duration)

Difference	Period	Duration
-1 hours	2018-03-11 03:00:00	2018-03-11 04:00:00

#Exit Daylight Savings
daylight_savings_end <- ymd_hm("2018-11-03 24:00", tz = "America/New_York")

period <- daylight_savings_end + hours(3)

duration <- daylight_savings_end + dhours(3)

tibble(Difference = difftime(period, duration, units = "hours"),
       Period = period,
       Duration = duration)

Difference	Period	Duration
1 hours	2018-11-04 03:00:00	2018-11-04 02:00:00

#Normal Year
normal_year <- ymd("2018-09-20", tz = "America/New_York")

period <- normal_year + years(1)

duration <- normal_year + years(1)

tibble(Difference = difftime(period, duration, units = "days"),
       Period = period,
       Duration = duration)

Difference	Period	Duration
0 days	2019-09-20	2019-09-20

#Leap Year
leap_year <- ymd("2019-09-20", tz = "America/New_York")

period <- leap_year + years(1)

duration <- leap_year + dyears(1)

tibble(Difference = difftime(period, duration, units = "days"),
       Period = period,
       Duration = duration)

Difference	Period	Duration
0.75 days	2020-09-20	2020-09-19 06:00:00

For months, imaginary dates (e.g. Feb 31) need to be dealt with:

ymd("2019-1-31") + months(1)

## [1] NA

ymd("2019-1-31") %>% add_with_rollback(months(1))

## [1] "2019-02-28"

ymd("2019-1-31") %>% add_with_rollback(months(1), roll_to_first = TRUE)

## [1] "2019-03-01"

ymd("2019-1-31") + dmonths(1)

## [1] "2019-03-02 10:30:00 UTC"

\(\color{skyblue}{\textrm{- Difference Between Times}}\)

difftime() measures the duration between two real datetimes

secs
mins
hours
days
weeks

#Normal Day
difftime(ymd_hm("2018-01-02 03:00", tz = "America/New_York"),
         ymd_hm("2018-01-01 24:00", tz = "America/New_York"), 
         unit = "days")

## Time difference of 0.125 days

#Enter Daylight Savings
difftime(ymd_hm("2018-03-11 03:00", tz = "America/New_York"),
         ymd_hm("2018-03-10 24:00", tz = "America/New_York"), 
         unit = "days")

## Time difference of 0.08333333 days

#Exit Daylight Savings
difftime(ymd_hm("2018-11-04 03:00", tz = "America/New_York"),
         ymd_hm("2018-11-03 24:00", tz = "America/New_York"), 
         unit = "days")

## Time difference of 0.1666667 days

\(\color{skyblue}{\textrm{- Intervals}}\)

interval() creates an interval

ryan <- interval(ymd("1990-11-04"), today())
ryan

## [1] 1990-11-04 UTC--2022-02-10 UTC

int_length() measures the length

int_length(ryan) %>% dseconds()

## [1] "986774400s (~31.27 years)"

int_shift() modifies the interval

int_shift(ryan, days(5))

## [1] 1990-11-09 UTC--2022-02-15 UTC

%within% determines if a date is within the interval

the_90s <- interval(ymd("1990-01-01"), ymd("1999-12-31"))

today() %within% the_90s

## [1] FALSE

int_overlaps() determines if two intervals overlap

the_1900s <- interval(ymd("1900-01-01"), ymd("1999-12-31"))

the_90s %within% the_1900s

## [1] TRUE

the_1900s %within% the_90s

## [1] FALSE

int_overlaps(the_90s, the_1900s)

## [1] TRUE

\(\color{dodgerblue}{\textbf{Combining Data}}\)

\(\color{skyblue}{\textrm{- Joins}}\)

inner_join()
full_join()
left_join()
right_join()

# Basic
joined_data <- data1 %>%
  inner_join(data2, by = "common_col")

# Multiple keys 
joined_data <- data1 %>%
  inner_join(data2, by = c("common_col1", "common_col2")

# Key with different name
joined_data <- data1 %>%
  inner_join(data2, by = c("col from df1" = "col from df2"))

\(\color{skyblue}{\textrm{- Filter Joins}}\)

sec_council <- tibble(country = c("United States", "United Kingdom", "France", 
                                  "Russia", "China"))

semi_join() returns data from one dataset that is present in a second dataset

gapminder %>% 
  filter(year == 2007) %>% 
  semi_join(sec_council, by = "country")

country	continent	year	lifeExp	pop	gdpPercap
China	Asia	2007	72.961	1318683096	4959.115
France	Europe	2007	80.657	61083916	30470.017
United Kingdom	Europe	2007	79.425	60776238	33203.261
United States	Americas	2007	78.242	301139947	42951.653

anti_join() returns data from one dataset that is NOT present in a second dataset

gapminder %>% 
  filter(year == 2007) %>% 
  anti_join(sec_council, by = "country") %>% 
  head()

country	continent	year	lifeExp	pop	gdpPercap
Afghanistan	Asia	2007	43.828	31889923	974.5803
Albania	Europe	2007	76.423	3600523	5937.0295
Algeria	Africa	2007	72.301	33333216	6223.3675
Angola	Africa	2007	42.731	12420476	4797.2313
Argentina	Americas	2007	75.320	40301927	12779.3796
Australia	Oceania	2007	81.235	20434176	34435.3674

\(\color{skyblue}{\textrm{- Binding}}\)

df1 <- tibble("A" = c(1, 2, 3),
              "B" = c(6, 5, 4))

df2 <- tibble("A" = c(4, 5, 6),
              "B" = c(3, 2, 1),
              "C" = c("L", "M", "N"))

df3 <- tibble("B" = c(4, 5, 6),
              "C" = c("1", "2", "3"))

bind_rows() tacks on new rows

df1 %>% 
  bind_rows(df2)

A	B	C
1	6	NA
2	5	NA
3	4	NA
4	3	L
5	2	M
6	1	N

bind_cols() tacks on new columns

df1 %>% 
  bind_cols(df3)

A	B…2	B…3	C
1	6	4	1
2	5	5	2
3	4	6	3

\(\color{skyblue}{\textrm{- Sets}}\)

country_z <- gapminder %>% 
  filter(year == 2007,
         str_detect(str_to_lower(country), "z"))

country_w <- gapminder %>% 
  filter(year == 2007,
         str_detect(str_to_lower(country), "w"))

union() returns all, removing duplicates

country_z %>% 
  union(country_w) %>% 
  head()

country	continent	year	lifeExp	pop	gdpPercap
Bosnia and Herzegovina	Europe	2007	74.852	4552198	7446.2988
Brazil	Americas	2007	72.390	190010647	9065.8008
Czech Republic	Europe	2007	76.486	10228744	22833.3085
Mozambique	Africa	2007	42.082	19951656	823.6856
New Zealand	Oceania	2007	80.204	4115771	25185.0091
Swaziland	Africa	2007	39.613	1133066	4513.4806

intersect() returns only observations in both

country_z %>% 
  intersect(country_w) %>% 
  head()

country	continent	year	lifeExp	pop	gdpPercap
New Zealand	Oceania	2007	80.204	4115771	25185.0091
Swaziland	Africa	2007	39.613	1133066	4513.4806
Switzerland	Europe	2007	81.701	7554661	37506.4191
West Bank and Gaza	Asia	2007	73.422	4018332	3025.3498
Zimbabwe	Africa	2007	43.487	12311143	469.7093

setdiff() returns only observations in the first that are not in the second

country_z %>%
  setdiff(country_w) %>% 
  head()

country	continent	year	lifeExp	pop	gdpPercap
Bosnia and Herzegovina	Europe	2007	74.852	4552198	7446.2988
Brazil	Americas	2007	72.390	190010647	9065.8008
Czech Republic	Europe	2007	76.486	10228744	22833.3085
Mozambique	Africa	2007	42.082	19951656	823.6856
Tanzania	Africa	2007	52.517	38139640	1107.4822
Venezuela	Americas	2007	73.747	26084662	11415.8057

\(\color{darkblue}{\textbf{Visualize}}\)

See ggplot or Spatial R

\(\color{darkblue}{\textbf{Model}}\)

See Statistics or tidymodels

\(\color{darkblue}{\textbf{Communicate}}\)

See Markdown or Shiny